RE: Draft-chiba-11 comments.

[Bernard Aboba <aboba@internaut.com>]

OK. I've added the requested attributes to the attribute table. Given
the size of the table, I've added a statement about
attribute changes being atomic -- either they are all made successfully,
or the Request MUST fail, with no changes.

Please check the table in the below draft very carefully from the original
source documents [RFC2865], [RFC2868], [RFC2869], [RFC3162]. We need to
be clear on the meaning of each attribute included for both CoA and
Disconnect cases. If in doubt, it would be best not to include an
attribute, rather than include it now and have interoperability problems
later. Here is the strawman draft:

http://www.drizzle.com/~aboba/IEEE/draft-chiba-radius-dynamic-authorization-12.txt

With respect to a request for re-authorization, my understanding was that this
draft was largely an attempt to document an existing protocol, and so that
adding major new features was not within scope.